Log inSubmit server
Sdd MCP logo

Sdd MCP

MCP tools for managing spec-driven development workflows, including specs, plans, approval gates, status, and session logs.

InstallRepositoryWebsite
Full README Compare alternatives Embed widget Claim this listing

Install Sdd MCP

Pick your client, then add it in one click or paste the config.

STDIONode.js
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "sdd-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@juanklagos/create-sdd-project@latest"
      ]
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

We couldn’t confirm it starts correctly

We ran the install command but it didn't respond within our test window — this can mean a slow first-time install rather than a real problem.

npx -y @juanklagos/create-sdd-project@latest

No response to initialize.

This is an experimental automated check and can have false negatives — missing environment variables, a slow cold install, etc. It doesn’t necessarily mean something’s wrong. Last checked 3d ago.

Overview

The sdd-mcp MCP server exposes a spec-driven development workflow for creating and managing specifications, plans, approval gates, status, and logbook records. It connects an AI agent to the repository’s SDD artifacts and validation process, helping keep implementation behind an approved spec and consistent plan. Use it when an existing or new project needs documented decisions, explicit consent, and repeatable workflow checks. The project supports English and Spanish guidance.

Use cases

  • Create approved feature specifications before implementation
  • Check whether a plan matches its specification
  • Record user consent and development decisions
  • Track SDD workflow status and next steps
  • Preserve session history in a project logbook

Key features

  • Specification management
  • Plan consistency checks
  • Approval gate workflow
  • Status and next-step tracking
  • Decision and logbook records
  • English and Spanish guidance

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked 2d ago

Recent health check succeeded.

6/6 checks healthy over the last 46d

Supply-chain signal

No high-severity advisories surfaced by our automated scan.

Critical 0High 0Medium 0Low 0

Scanned 1d ago via OSV.dev · @juanklagos/create-sdd-project@latest (npm)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

GitHub stars21Stargazers on the source repository.
Last commit1mo agoMost recent push to the default branch.
Install checkInconclusiveDidn't respond in our test window — often a slow first install.

How Sdd MCP works

What sdd-mcp does

The sdd-mcp MCP server brings the repository’s Spec-Driven Development workflow into an AI agent through MCP tools. Its focus is the work that happens before and around implementation: defining a feature, preparing a plan, checking approval conditions, tracking progress, and recording decisions or session history.

The workflow treats written artifacts as the source of project decisions. A feature is organized into a numbered bundle containing a specification, plan, tasks, and history. The repository also uses a bitacora/ logbook to preserve session context and decisions for later review.

The central policy is that implementation should not begin until an approved spec.md and a consistent plan.md are present, with user consent recorded. This makes the server relevant for teams using AI coding agents that need a clear boundary between requirements and code changes.

How it works

The workflow starts with an idea expressed in plain language. The agent then helps create or refine a specification, including acceptance criteria, and prepares a plan that matches the approved scope. Tasks can be prioritized before the workflow reaches its gate.

At the gate, the project checks approval, plan consistency, and recorded consent. The repository provides shell scripts for policy checks, gate validation, and consent recording; the MCP connector allows an AI tool to operate the workflow rather than relying only on manual commands.

After implementation, validation and logbook updates complete the cycle. Decisions can be written to bitacora/decisiones/, including the decision, rationale, rejected alternatives, and a possible revisit point.

Tools and capabilities

  • Creating and managing specifications.
  • Preparing and reviewing implementation plans.
  • Running or working with approval gates.
  • Checking workflow status and identifying the next step.
  • Recording decisions and session history in the logbook.
  • Supporting the broader SDD process in English and Spanish.

The repository also includes related agent workflows outside MCP. Claude Code commands cover help, new specifications, spec refinement, gate execution, decisions, session closure, and tutoring. VS Code and Copilot use prompt files, while a portable Agent Skill supports any agent that can consume it. These companion entry points are part of the repository’s wider workflow and should not be confused with the MCP tool surface.

Setup and configuration

The repository describes three ways to apply the workflow to a project: a compact spec/ sidecar for an existing codebase, an internal workspace, or a full standalone copy. A project can also be scaffolded with npx @juanklagos/create-sdd-project@latest, which asks configuration questions and creates the selected project structure.

For an existing project, the sidecar approach keeps application code in its current location while placing SDD artifacts under ./spec/. The repository also documents shell scripts for installing a sidecar, creating a workspace, initializing a project, checking policy alignment, validating the SDD gate, and recording consent.

Limitations and notes

The sdd-mcp MCP server is tied to the repository’s documented SDD conventions, including numbered spec bundles, approval records, plan consistency, and the bitacora/ logbook. It is best suited to projects willing to keep these artifacts under version control.

Exact tool names and client configuration therefore cannot be inferred from this listing alone.

The repository is MIT licensed and includes bilingual English and Spanish learning material. Its broader template also builds on GitHub Spec Kit for step-by-step commands and workflow structure.

Read the full README View source on GitHub

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about Sdd MCP

sdd-mcp is an MCP server that connects AI agents to the Spec-Driven Development workflow in the spec-driven-development-template repository. Its main tools manage specifications, plans, approval gates, workflow status, and logbook records, helping teams require an approved spec and consistent plan before implementation begins.

It documents project setup through a spec sidecar, internal workspace, or standalone copy, plus an npx scaffolding command, but the connector’s exact installation steps must be checked in the repository.

Cursor is named as an AI assistant that can use the repository’s SDD workflow.

It supports a workflow gate that requires an approved spec, a consistent plan, and recorded consent before implementation.

Related MCP servers

All in Product Management All alternatives
Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewSdd MCP AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/sdd-mcp?style=directory)](https://allmcps.com/mcp/sdd-mcp)
HTML Embed
<a href="https://allmcps.com/mcp/sdd-mcp"><img src="https://allmcps.com/api/badge/sdd-mcp?style=directory" alt="Sdd MCP on AllMCPs" /></a>